Organization-wide approaches. In Latin America top leaders in two countries launched large-scale quality improvement initiatives in 1991. Chile's National Program for Evaluation and Improvement of Quality (EMC) is a Ministry of Health initiative that has been phased into all of the nation's health regions with assistance from the Quality Assurance Project (QAP) (87). EMC is now an intrinsic part of regional health management. In Mexico MEXFAM, an IPPF affiliate, has moved more slowly. It first introduced CQI in 7 of its 36 logistic regions and then recently extended the program to additional cities (266, 377).

Both organizations made significant investments in quality improvement, especially in training, creating new organizational structures to oversee quality improvement activities, and revamping monitoring systems. Chile spent US$500,000 on quality improvement over two and a half years, largely to train more than 5,200 health workers and 250 quality monitors and to create 90 quality committees and commissions (35, 118). The government health system covers continuing costs (87, 118).

MEXFAM's initiative, began with an investment of US$112,000, is less well institutionalized than Peru's (35). An evaluation found that QI teams continue to operate and often provide valuable solutions, but many have not mastered the methodology, and there has been almost no effort to replicate or expand successful innovations beyond their original sites (377).

Grass-roots approaches. Other QI approaches operate in individual facilities with minimal investment of time and money. Small-scale initiatives may work better and cost less than centralized efforts that require extensive training. Thus recent QAP programs have used data available locally rather than redesign central monitoring systems and have relied on training from local mentors (300).

COPE is the most widely used grass-roots approach. Outside facilitators launch the COPE process, but staff teams are supposed to take charge after the first year and continue without further outside support (23, 219). Evaluation of COPE exercises at four African clinics, however, found that clinic teams did not feel ready to carry on alone after a year, and only one clinic showed clear signs of integrating COPE into its management structure (242). At the Family Planning Association of Kenya (FPAK), which helped develop COPE in the late 1980s and has the greatest experience with the approach, staff members have mastered the method, however. Each FPAK clinic conducts a COPE exercise every three months (44).

It is unclear how much impact a grass-roots approach such as COPE can have on a large program. Some nationwide family planning organizations, such as FPAK, have trained a corps of COPE facilitators to introduce the technique throughout their entire network of clinics (35, 219). The next step is to move COPE from the clinics to higher management levels. COPE tools may not be appropriate for central management functions, however (44).